Problem: can't open device /dev/ttyMH. **solved**

[kusti]

Dear all

I'm operating a Meteobridge PRO with a Davis 6322 since July 2019 providing meteo data for my webpage (https://www.kusti.ch/wetterstation/). Everything ran quite well until I had to disconnect my Meteobridge PRO from power for a few days, since then I'm not able to receive any data of my Davis station, looks like the firmware is not able to connect to the internal receiver (just a guess!):

Code: Select all

logger (31.12.2022 13:41:14): data logger (version 5.4, build 14873) started.
logger (31.12.2022 13:41:15): station 0 (Meteostick Vantage), can't open device /dev/ttyMH: No such file or directory
logger (31.12.2022 13:41:15): station 0 (Meteostick Vantage), connect to station failed.
logger (31.12.2022 13:41:15): connect station 9 (System Data via Plug-in).
logger (31.12.2022 13:41:18): waiting for sensors (wind0, rain0, th0, thb0) to provide data.
logger (31.12.2022 13:41:18): services paused.
logger (31.12.2022 13:41:45): station 0 (Meteostick Vantage), can't open device /dev/ttyMH: No such file or directory
logger (31.12.2022 13:42:45): message above repeats 2 times.
database (31.12.2022 13:43:02): database sync done (17 tables, 170/170 writes).
logger (31.12.2022 13:43:15): give up waiting for sensors after 120 secs.
logger (31.12.2022 13:43:15): services activated.
logger (31.12.2022 13:43:15): station 0 (Meteostick Vantage), can't open device /dev/ttyMH: No such file or directory
logger (31.12.2022 14:08:15): message above repeats 50 times.

I was using a quite old firmware (unfortunately didn't write it down, but was from 2019) and so I hoped updating the system would solve the problem. I also tried to reconfigure the weather station, without any success. I can't select a device, the dropdown menu is empty (https://drive.google.com/file/d/1plXI55 ... sp=sharing)

I found an entry in the forum here (viewtopic.php?p=41392&hilit=%2Fdev%2FttyMH#p41392) but looks like there was no happy end (at least not public in the thread).

Is there any chance to fix this by software / configuration or is this really a hardware failure?

Thanks for any suggestions!

Christian

[WetterCH]

Hallo Christian

Schau mal das Bild an, hast du diesen Knopf bei dir „RF-Modul“ Firmware-Update?

Eventuell kannst du damit eine neue FW draufspitzen, und es so wieder zum laufen bringen.

Nachtrag:

Versuche auch mal noch eine ältere Firmware auf der MB-Pro zu installieren, bei mir gehts bis 3.2 zurück.
Sofern du keine Verbindung zum RF-Empfänger hast.

Es kann gut möglich sein, dass die neuste Firmware auch ein Update des RF-Moduls erfordert.
Viel Erfolg beim ausprobieren.


Gruess Dominic

[kusti]

Hallo Dominic

Danke für den Tipp! Leider kann mit dem Firmware Update das Problem nicht gelöst werden, es scheint so, wie wenn er den internen Funkempfänger nicht mehr erkennt:

Bildschirm­foto 2023-01-02 um 16.44.29.png (71.54 KiB) Viewed 3063 times

Das Problem kam von einem Tag auf den anderen, ohne Firmware Update. Ich habe daraufhin die Firmware auf die neuste Version aktualisiert und umgekehrt auch wieder auf einen alten Stand gesetzt, alles leider ohne Erfolg...

Ich habe mir nun einen WDI Meteostick bestellt, entweder funktioniert dieser als externer Empfänger an der Meteobridge oder ich schreibe mir dann selbst ein Script für meinen Raspi.

Gruss
Christian

[seatonsurfer]

I had exactly the same issue just before Christmas and Boris logged in and concluded that it was a usb hub failure on the motherboard. I suggest you contact Boris and ask him to have a look at your Meteobridge Pro.

Regards

Neil

[kusti]

Hi Neil

Thanks for your answer! Leads to two questions:

First, who is Boris (I'm a newbie in this forum...)? And what was the solution of the problem, respectively, was there a fix without replacing hardware?

Regards
Christian

[BigAlT]

Boris is the developer of Meteobridge and the administrator of this Forum. His email address is info@meteobridge.com. He will probably ask you to give him remote access to your Meteobridge to diagnose the problem if it is accessible from the web. If not, it is probably a hardware issue.

[kusti]

Thanks, I'm going to write him:

Received the meteostick and getting now data, but the protocol has still quite a lot of errors in it...

[kusti]

Closing this issue now, looks like my meteobridge is broken:

Boris found no USB device connected, neither the internal USB stick nor the RF interface. I opened the device then and found a solid working USB memory stick inside, I guess the USB part of the uC inside the meteobridge must be broken

As the meteostick is too expensive to replace (and offers much more functionality than I need for my purpose) I decided to write my own node.js application on an already existing Raspberry Pi 2 (with Ubuntu server) connecting to a meteostick.

Thanks for the support!
Christian

[Lukas]

Hi! Well, you know, my Meteobridge has been giving the same error since this morning! Is it just a coincidence?

[kusti]

Lukas, do you still have the internal memory or is it inaccessible too?

[Mattk]

Hi! Well, you know, my Meteobridge has been giving the same error since this morning! Is it just a coincidence?

How about you post some screen shots of error you are seeing on your unit?

[Lukas]

Sure, thank you for asking! The log file looks like this:

Code: Select all

logger (11.01.2023 09:15:05): station 0 (Meteostick Vantage), can't open device /dev/ttyMH: No such file or directory
database (11.01.2023 09:15:05): cannot prepare select on update table 'data10num_hour' (after 0 tries in 1ms): no such table: data10num_hour
database (11.01.2023 09:15:05): rollback on insert forced (2) (hour): data10num
database (11.01.2023 09:15:05): cannot create table (after 0 tries in 1ms): database or disk is full (13)
database (11.01.2023 09:15:05): cannot commit (after 0 tries in 1ms): cannot commit - no transaction is active (1)
database (11.01.2023 09:15:05): cannot rollback (after 0 tries in 1ms): cannot rollback - no transaction is active (1)
database (11.01.2023 09:15:05): cannot prepare select on update table 'data10num_hour' (after 0 tries in 1ms): no such table: data10num_hour
database (11.01.2023 09:15:05): rollback on insert forced (2) (hour): data10num
database (11.01.2023 09:15:05): cannot create table (after 0 tries in 1ms): database or disk is full (13)
database (11.01.2023 09:15:05): cannot commit (after 0 tries in 1ms): cannot commit - no transaction is active (1)
database (11.01.2023 09:15:05): cannot rollback (after 0 tries in 1ms): cannot rollback - no transaction is active (1)
database (11.01.2023 09:15:05): cannot prepare select on update table 'data10num_hour' (after 0 tries in 1ms): no such table: data10num_hour
database (11.01.2023 09:15:05): rollback on insert forced (2) (hour): data10num

And this same block repeats over, until it reaches the last line, which is truncated and looks like this:

Code: Select all

database (11.01.2023 09:16:02): cannot pr

The webcam image is showing, but the the meteo station data is not:

camplus.jpg (68.44 KiB) Viewed 2898 times

It all started yesterday at around 8.40 am, which is when Windfinder received the last update. Since then I tired rebooting, powering down, downgrading to the previous firmware version (you never know), going back to the latest firmware 5.5, and updating the RF Module firmware, which gives the same error that Kusti has:

Immagine 2023-01-11 093510.jpg (34.5 KiB) Viewed 2898 times

If I can provide any other detail please do ask, and thank you for your help!!

[kusti]

Hi Lukas

First of all, nice view of the webcam!

The logs provided are the same as mine did after I tried to use the internal memory: you don't have a connection to the /dev/ttyMH device AND the internal disk is inaccessible, considered as full.

As far as I have an understanding of the electronical parts in the Meteobridge I assume that the USB part of the microcontroller is broken and therefore the software is not able to connect to the USB memory and RF part as well. I rule out a software problem, otherwise a downgrade should have solved the problem or at least Boris would have recognised this.

In this forum alone, that's two units within two weeks with the same error pattern, which makes me wonder if it's a coincidence or a systematic problem.

Regards
Christian

[admin]

Remote login reveals that there is a HW defect that needs repair.